A JOURNAL ON TAXONOMIC BOTANY, PLANT -SOCIOLOGY AND ECOLOGY REINWARDTIA . Editors MIEN A. KIJS WAT A KARTAWINATA N. WULIJARNI-SOETJIPTO 1 Published by ' HERBARIUM BOGORIENSE LEMBAGA BIO-LOGI NASIONAL — LIPI BO.SOR, INDONESIA Eeinwardtia Vol. 9, Part 1, 1 —182 31 December 1974 10- ISSN 0(f34-365X REINWARDTIA Published by Herbarium Bogorlense — LBN, Host Vol. 9, Part 1, pp. 77 — 83 (1974) REDISCOVERY OF CHEILOTHECA MALAYANA AND THE IDENTITY OF CHEILOTHECA, ANDRESIA AND MONOTROPASTRUM (ERICACEAE-MONOTROPOTDEAE) HSUAN KENG Depavlrnvni of Botany Univgrsiti/ of S7 ABSTRACT Andreeia, Monotropastri my of Ckeilotheca. Four spe a new combination, Ckeitothe species, Chtiilothtfcti si? apore, Singapore a ted; Andr nd Wirtge-nia are reduced to the are accepted, keyed out ami enum en huiailie (D. Don) H. Keng, and a new H. Rene, are proposed. ABSTKAK notropastruM dan Wirtgenia diperialmkan sebairai ainonim Ckeilotkeea. Dalam marga ini diterima adanya empat jenis yang dapat dibeda-bedatan berdasarkan suatu kunci detepminaai. Satu kom- binasi baru Cheiloiheea humilis (D. Don) H. Kene dan aatu jenis baru Cheilotheaa aleumertima H. Keng telah diusulkan. In late October 1972, my colleagues, Drs. C. J. Goh, K. H. Chow and I, together with the Honours students of the Botany Department, University of Singapore, went on a week-long field trip to Maxwell Hill, Taiping, Perak, W. Malaysia. One day", on October 24th, we went along a small foot path off the 5*/> milestone of the main road. This led us into a thick wood In the belt of the upper dipterocarp forest. The altitude is around 2900 ft. or 870 m. On the dark, damp forest floor, among other things, we collected ft small, creamy white-coloured plant whose black, fibrous roots were extensively growing on rotting leaf litter (Figs. A, B and C). We were somewhat puzzled by the showy, yellow, terminal, 3-merous flowers. Only after we returned to Singapore was it identified by Professor R. E. Holttum as Cheilotheca malaya.no, Scort. ex Hook. f. The genus Cheilotheca. was first established by J. D. Hooker in 1876 based on a species collected by himself from Khasi Hills of Eastern India. The Malayan species is the second species of the genus, it was discovered and described (in manuscript) by Father B. Scortechini (and later also collected by H. Kunstler) from Larut, Perak, in the 1880s. These collections are presently preserved at Kew. This plant has never been collected afrain since and was not even included in Burkill and Henderson's local flora of Taiping {in Gard. Bull. Str. Settlem. 3, 1925). —77 — REINWARDTIA [VOL, 9 1974] KENG: Ckeilothi Rede fined This curious Malayan plant, Cheilotheca malayana, was later segregated into a monotypic genus Wirtgenia by H. Andres (1914) which is unfortunately a later homonym of at least 3 entirely different genera, among them the earliest one appears to be Wirtgenia Sch. Bip. (1842), a genus now known as Aapilia, Thou. (Compositae). For this reason, H. Sletimer, in his treatment of Ericaceae for the Flora Malesiana (1967), renamed it as Andresia. Dr. Sleumer who was handicapped by lack of suitable preserved material, therefore tentatively accepted Andres' view and. pointed out that the Malayan plant cannot be congeneric with Gheilotlieca on the grounds tabulated below. Cheilotiwea (Indian) 1. Sepaloid scales 3 - 4 only; 2. Petala flat; 3. Anthers of parallel cells with con- tractions or swellings at jntervals; each cell opening hy an irregular longitudinal slit; 4. Stigmas in the form of a pileus or reversed cup; 5. Placentas laterally expanded and bear few ovule? only. Andreda (Malayan) !. Sepaloid scales 5; 2. Petals concave; 8. Anthers linear, basifixod, hippocre- piforra; cells 2, smooth, confluent at apex, dehiscing by marginal pores which gradually extent into slits; 4. Stigmas obscurely 4-lobed or 8-lobed; 5. Pla 6, parietal, on promi slightly i merous Plate 1 CheiUHkecz • " * • » " " • : Hook. f. (Divisions in 1 mm). A and B. D.T-Eruiting plant. E. fruit seen from above. These five points of distinctions are briefly discussed as follows: (1) The sepaloid scales are transitional between petals (or petaloid seales) and scale-leaves (Figs. F, H and I). From the Malayan material studied, their numbers are mostly 3, rarely 2 (Fig-. I), but never 5 as stated by Ridley (1923). (2) The petals of the Malayan plant are clearly concave. No Khasian material is available for this study. But Hooker's original description stated (p. 607) that "petala . . . non saccata, imbricata". As they are imbricate and their number being 3, it is thus rather unlikely that they are flat. (3) This distinction more or less stands (for details, see key below). (4) From the Malayan material studied, the stigma is hemispheric, but shallowly depressed in the centre, and very obscurely 6-lobed (PiEs. G and H ) ; in the fruiting stage, however, it becomes cup-shaped (FiK- E ) . This is generally in agreement with the description of the Khasian plant. REINWARDTIA [VOI. 1074] KENG: Cheilotheca redefined (5) Serial microtome sections of the ovary of the Malayan specimen show different configurations at different levels. In the sections around the middle portion, the placentas are less expanded and bear numerous ovules; whereas in the sections near the base, the placentas are strongly centripetally expanded and bear fewer ovules. Hooker in his original description of the Khasian plants stated that "ovula in loculis numerosis- sima". Therefore the only remaining- substantial difference between the Indian and Malayan plants is the external morphology of their anthers, as clearly pointed out by Hooker earlier (1887). It appears reasonable therefore to retain the Malayan plant within the genus Cheilotheca. In reviewing the literature on this subject, my attention was drawn to another closely related genus, Monotropastrum, which was also created by H. Andres (1935). Originally there were three species described under this genus, namely, M. macrocarpum Andres (from S.W. China and B. India), M. ampullac&us Andres (S.W. China) and M. arixanarum. Andres (Formosa). Of these three, I have only a faint recollection of the last species which I observed and collected many years ago. Professor H. Hara (1941, 1961, 1965) has made a series of studies of this genus. He examined both the fresh and herbarium materials of the Himalayan and the Japanese plants (the latter was excellently illustrated in T. Nakai, Icon. PI. Asia. Orient. 4 ( 1 ) : 324. 1941), and finally came to the conclusion (1965, 1972) that there is only one polymorphic species, namely, Monotropastrum humile (D. Don) Hara, in the genus. He also compared this genus with Monotr&pa (especially M. nviflora L.) and concluded that Monotropastrum differs from the latter in the following characters: (1) smooth unilocular ovary with 6—13 parietal placentas, (2) anther opening by an elliptic lid, (3) generally bluish stigmas, (4) indehiscent berry and (5) oval seeds without appendages. The fruit of the Khasian plant was at first unknown, but later Hooker (1887) was informed by C. B. Clarke that it was a subglobose berry. The material we recently collected from Maxwell Hill, Perak also included a leathery berry (Figs. D and E) with oval seeds devoid of appendages. The tangible difference between Cheilotheca and Monotro- pastrum, like the difference between Cheilotheca and Andre&ia, lies solely on their androecial characters. I therefore suggest to reinstate the Malayan species to Cheilotheca and to reduce Monotropastrum Andres to a synonym of Cheilotheca. Also based on the descriptions and illustration provided by Sleumer (1967) and on phytogoographieal ground, I would like to consider the Sumatran 1..., Scort. ex Hook. f. (Divisions in 1 cm) P dissection ring (1) andro-gynoecium (centre), (2) 3 peta'loid scales, (4) an uppermost leafy 3 C B ] e . G. androgvnnedun, of p - (different from F> with the S stamens showing (1) andro-gynoesium, (4) an uppermost leafy scale, nlarged. 2 sepaloid stales, am andro-gynoecium of I, 7